Elephant in the Fridge: Constant-Memory Frame Packing for Long Video Understanding
Paper • 2610.16118 • Published • 22
Zhi Wei Tan
zwtan-cs
AI & ML interests
#distillation, model-based RL
Recent Activity
liked a dataset 1 day ago
emgena/speculative_decoding_draft_guard_teaser upvoted a paper 3 days ago
Contextual Correction of Parallel Drafts for Speculative DecodingOrganizations
None yet
upvoted a paper 1 day ago
upvoted 3
papers 3 days ago
Contextual Correction of Parallel Drafts for Speculative Decoding
Paper • 2610.12570 • Published • 35
Surfing the Activation Wave: Analyzing the Information and Geometry of LLM Hidden States for Speculative Decoding
Paper • 2610.12939 • Published • 17
One Model to Loop Them All: Elastic Latent Recurrence for Language Models
Paper • 2610.12582 • Published • 15
upvoted a paper 4 days ago
upvoted 3
papers 7 days ago
HiPLEX: Hierarchical Policy Factorization for Full Duplex Speech Language Models
Paper • 2610.07727 • Published • 17
RealtimeWAM: One-Step Asynchronous World Action Models
Paper • 2610.06617 • Published • 24
QuantWM: Temporally Consistent 2-Bit KV Cache Quantization for Video World Models
Paper • 2609.26425 • Published • 23
upvoted 7
papers 9 days ago
Selection-Based Structured Reasoning: Toward Efficient Multimodal Search Agents
Paper • 2610.01892 • Published • 38
SlimWise: Decoupling Expert Pruning Across Prefill and Decode for Efficient MoE Serving
Paper • 2609.34117 • Published • 25
Learning Functional Subspaces for Neural Network Compression
Paper • 2609.40127 • Published • 25
Making LLMs Say What They Think: Measuring and Improving CoT-Interpretability Alignment
Paper • 2609.38972 • Published • 29
HLA-WM: Hybrid Linear Attention for Long-Horizon Video World Models
Paper • 2610.05739 • Published • 21
Towards Looped Models Done Right, Part II: Rethinking at Fixed Points
Paper • 2610.06833 • Published • 35
ProgressCompass: Embodied Progress Reward Models Are Lost Without the Right Context
Paper • 2609.36684 • Published • 20
upvoted 4
papers 12 days ago
Prefill-Free Cross-Family KV Cache Transfer for Heterogeneous Multi-Agent LLMs
Paper • 2609.32259 • Published • 103
On-Policy or Off-Policy Learning? A Systematic Study of Distillation Dynamics
Paper • 2609.35259 • Published • 205
WUSH-KV: KV Cache Quantization with Data-Adaptive Transforms
Paper • 2609.38121 • Published • 27
OSWorld-Science: A Benchmark of Computer Use Agents for Learning and Using Scientific Software
Paper • 2609.39903 • Published • 64
upvoted a paper 14 days ago